CN103034838B - A kind of special vehicle instrument type identification based on characteristics of image and scaling method - Google Patents

A kind of special vehicle instrument type identification based on characteristics of image and scaling method Download PDF

Info

Publication number
CN103034838B
CN103034838B CN201210509245.8A CN201210509245A CN103034838B CN 103034838 B CN103034838 B CN 103034838B CN 201210509245 A CN201210509245 A CN 201210509245A CN 103034838 B CN103034838 B CN 103034838B
Authority
CN
China
Prior art keywords
instrument
image
special vehicle
type
normalization
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN201210509245.8A
Other languages
Chinese (zh)
Other versions
CN103034838A (en
Inventor
张伏龙
白真龙
王涛
郭迎春
张华�
李广峰
赵玉林
张昌俊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
63963 TROOPS PLA
University of Science and Technology Beijing USTB
Original Assignee
63963 TROOPS PLA
University of Science and Technology Beijing USTB
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 63963 TROOPS PLA, University of Science and Technology Beijing USTB filed Critical 63963 TROOPS PLA
Priority to CN201210509245.8A priority Critical patent/CN103034838B/en
Publication of CN103034838A publication Critical patent/CN103034838A/en
Application granted granted Critical
Publication of CN103034838B publication Critical patent/CN103034838B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Image Analysis (AREA)

Abstract

A kind of special vehicle instrument type identification based on characteristics of image and scaling method.The invention belongs to image procossing, area of pattern recognition.First to the special vehicle instrument of every type, gather some special vehicle instrument images, through artificial judgment, representational image is stayed and is used as training sample; For each training sample, by image pre-processing method, picture quality is normalized; Then extract the disk of Instrument image, according to disc radius, size normalization is carried out to Instrument image; Respectively to special vehicle instrument image zooming-out color characteristic and Gabor textural characteristics after normalization; Last for all training samples, respectively with color characteristic and Gabor textural characteristics for vector is every type Modling model; After the characteristic model of the training sample database of the instrument of every type establishes, for the realtime graphic collected, picture quality normalization is carried out equally through Image semantic classification, respectively pattern match is carried out to the feature templates of the instrument training sample of these two kinds of features and each type, obtains instrument classification results by arest neighbors rule.

Description

A kind of special vehicle instrument type identification based on characteristics of image and scaling method
Technical field
The invention belongs to image procossing, area of pattern recognition, mainly for be the image of various dissimilar special vehicle instrument, the type of special vehicle instrument is identified automatically, and in conjunction with the prior demarcation of all types of instrument, provides the position of pointer point of fixity and meter dial.
Background technology
Automatic Measurement Technique has a wide range of applications in the production of special vehicle instrument, such as when special vehicle instrument needs automatically to detect.For a long time, the calibration of special vehicle instrument generally adopts manual read's access method with measurement.Although manual detection is accurate, there is very large inconvenient part.Carry out instrument calibration employee need manually check total indicator reading, with the comparison of standard source numerical data, the error of calculation, saving result, artificial operate miss and the collimation error of people can be there is when pointer quick rotation, and examine and determine efficiency can be very low.Therefore the calibrating of automatic instrument is in demand.The non-contact automatic detection technology of pointer instrument can be realized by image processing techniques, namely utilize imageing sensor to simulate human eye, image acquisition is carried out to pointer instrument, then by gather video data transmission in the intelligent system in computing machine.Recycling computing machine is analyzed the Instrument image collected and identifies, finally obtains the result pointed by gauge pointer.
In the past in the research of pointer instrument, mainly for the automatic identification of instrument, and mainly carry out studying for industrial instrument and compare, as document 1 (Chang Faliang, perhaps talented, Qiao Yizheng, the automatic identification and analysis method of unmanned indicator real-time vision, electronic surveying and instrument journal, 2006.4,20 (2): 35-38.), document 2(Sun Lin, Wang Yongdong, pointer instrument automatic Verification image recognition technology, modern electronic technology [J], 2011.Vol.34 (8): 101-104.) etc.For research lacking very of special vehicle instrument identification, less to the automatic identification of special vehicle instrument type.Special vehicle instrument is very different than industrial instrument, first special vehicle instrument pointer point of fixity is not often the central point of instrument disk, secondly the scale of very large special vehicle instrument is often uneven, brings very large difficulty also just to the automatic identification of pointer point of fixity and meter dial.If can for the modeling of particular meter type, so just automatically can identify the meter type of the Instrument image of input, according to the prior demarcation of the type, just can obtain the necessary information that the instrument such as the position of pointer point of fixity, the distribution of scale identifies automatically, thus be conducive to the automatic reading of special vehicle instrument.
Summary of the invention
Object of the present invention, be for be the image of various dissimilar special vehicle instrument, the type of special vehicle instrument is identified automatically, and in conjunction with the prior demarcation of all types of instrument, provides the position of pointer point of fixity and meter dial.
Technical scheme of the present invention is: identifying and modeling special vehicle instrument type based on characteristics of image.Extract the algorithm of feature and mainly combine two kinds of characteristics of image, one of them is color characteristic, and another one is Gabor textural characteristics, by these two kinds of characteristic synthetics, then to classify to special vehicle instrument type by the method for template matches and identifies.
First to the special vehicle instrument of every type, gather some special vehicle instrument images, through artificial judgment, representational image is stayed and is used as training sample.For each training sample, by image pre-processing methods such as image enhaucament, picture quality is normalized; Then extract the disk of Instrument image, according to disc radius, size normalization is carried out to Instrument image; Respectively to special vehicle instrument image zooming-out color characteristic and Gabor textural characteristics after normalization.Last for all training samples, respectively with color characteristic and Gabor textural characteristics for vector is every type Modling model.Its idiographic flow as shown in Figure 2.
After the characteristic model of the training sample database of the instrument of every type establishes, for the realtime graphic collected, picture quality normalization is carried out equally through Image semantic classification, extract instrument disk and carry out instrument size normalization, color characteristic and Gabor textural characteristics is extracted respectively after normalization, carry out pattern match respectively to the feature templates of the instrument training sample of these two kinds of features and each type, the special vehicle instrument arest neighbors rule for little type obtains instrument classification results.Its idiographic flow as shown in Figure 1.
Main research of the present invention has: (1) Instrument image normalization: comprise picture quality normalization, Instrument image size normalization; (2) feature extraction: comprise color feature extracted, Gabor texture feature extraction; (3) feature modeling and pattern match; (4) according to the demarcation of specific meter type, gauge pointer point of fixity, calibration information is obtained by meter type.
1, Instrument image normalization
Its fundamental purpose of the normalization of special vehicle instrument image reduces the variation in same meter type between different sample, namely strengthens the degree of polymerization in class.In the class of typical special vehicle instrument, variation comprises several below: the illumination of instrument dial plate, the picture quality of instrument, instrument dial plate size, etc.Make a variation in these classes, first will be normalized Instrument image quality, this respect mainly relies on pre-service to reduce noise circumstance difference and the photoenvironment difference of various image as far as possible.Then will carry out size normalization to Instrument image, this respect depends on the extraction of instrument disk and radius.Therefore the normalization of special vehicle instrument image includes several step, sequentially successively:
● Instrument image strengthens and removes denoising,
● instrument disk extracts,
● according to the size of disk to the linear normalization of instrument size
Carrying out pre-service to special vehicle instrument image, is exactly briefly to make the picture quality difference of image and the template image tested minimum.Pre-service can solve image because the Instrument image that causes of the reason such as light or shooting angle is fuzzy, crooked or defect abnormal conditions.Here take and go random noise by medium filtering, being stretched by color strengthens the contrast of Instrument image.Best with adaptive median filter effect in removal noise and preservation details two again in medium filtering, we adopt the modified adaptive median filter algorithm combining mean filter and adaptive median filter advantage.And the adaptive neighborhood method of average removal of images interference using band to revise and noise.
In order to the coloured image of instrument is drawn high.We first obtain gray level image corresponding to coloured image, and add up the minimum gray value g1 in Instrument image, gray scale maximal value g2, are drawn high the gray space of (0,255), make gray level image more clear.Here adopt a kind of gray scale of improvement to draw high method, considers minimum gray value g1, little to general image improved action time the grey scale pixel value number corresponding to gray scale maximal value g2 is little.The total number of the gray-value pixel that we require g1 and g2 corresponding is configured to need to be greater than certain value.If g1 is the minimal gray that total pixel is greater than n (such as n=10), g2 is the maximum gray scale that total pixel is greater than n (such as n=10).For any pixel, its gray-scale value is gray (x, y), and the gray-scale value drawing high rear correspondence is GRAY (x, y), then have:
GRAY ( x , y ) = 255 × [ gray ( x , y ) - g 1 ] ( g 2 - g 1 ) - - - ( 1 )
For the coloured image of RGB pattern, the redness of its correspondence, green, blue color draw high and are respectively:
R ( x , y ) = 255 × [ r ( x , y ) - g 1 ] ( g 2 - g 1 ) - - - ( 2 )
G ( x , y ) = 255 × [ g ( x , y ) - g 1 ] ( g 2 - g 1 ) - - - ( 3 )
B ( x , y ) = 255 × [ b ( x , y ) - g 1 ] ( g 2 - g 1 ) - - - ( 4 )
The method that we utilize Hough transform to detect circular arc obtains the edge of special vehicle instrument disk.After obtaining the size of instrument disk, we carry out linear normalization to the size of instrument.So-called size normalization be exactly test pattern and template image all scaling to identical size.Normalization fundamental purpose is the variation of radius size between image in order to reduce Real-time Collection and template instrument types of image.What we did is crop the region outside instrument disk, then the pixel in all instrument disks is normalized into the image that radius is 64 pixel sizes.During linear normalization, only do the conversion of equal proportion at X and Y direction.After normalization, Instrument image becomes the rectangle of 128x128 size, and the edge and then instrument disk of rectangular image, the pixel in the region outside instrument disk is all arranged to white pixel.
2, feature extraction
Color characteristic is one of key character of special vehicle instrument.Be different from industrial instrument, the local of some special vehicle instrument is colored, as the instrument in Fig. 3, Fig. 6, Fig. 9.Major part instrument is black matrix, and as the instrument of Fig. 3, Fig. 5, Fig. 6, Fig. 7, Fig. 8, Fig. 9, Figure 10, Figure 11, Figure 12, and the color of dissimilar special vehicle instrument is different.According to our observation, in special vehicle instrument, red, green, black is most important several color, and therefore we carry out feature extraction for these colors.The Instrument image that normalizing is good is divided into the grid of 8X8 by us, and each grid adds up total number that is red, green, black picture element respectively, then divided by the area of this grid, constitutes the color feature vector of 8X8X3=192 dimension like this.
For Instrument image based on Gabor gray level image texture feature extraction, we are for the gray level image of the good instrument of normalization, and each gray level image is divided into the grid of 8X8, gets each side's center of a lattice, this generates the sampled point of 8X8.On each sampled point, we add a Gaussian wave filter, and Gaussian wave filter is here tried to achieve from the Gaussian envelope of Gabor filter, and detailed process as the following formula.
G ( x , y ) = κ 2 σ 2 exp [ - κ 2 ( x 2 + y 2 ) 2 σ 2 ] = 4 λ 2 exp [ - 2 ( x 2 + y 2 ) λ 2 ] - - - ( 5 )
F d ( x i , y j ) = Σ x = - N x = N Σ y = - N y = N f d ( x i + x , y j + y ) G ( x , y ) - - - ( 6 )
Our setting parameter σ=π in the equation above, wherein wavelength X=8, N=2 λ.
Like this centered by each sampled point, be that non-zero point is sued for peace by Gaussian wave filter to pixel values all in Gaussian envelope, just on each sampled point of each template image, obtain an eigenwert, altogether obtain 8X8 (sampled point)=64 dimensional feature vector.Proper vector is by obtaining 64 final dimensional feature vectors to the equation of each eigenwert extraction of square root
3, feature modeling and pattern match
Try to achieve instrument template type image proper vector and by glyphomancy instrument image proper vector after, calculate the Euclidean distance d between them, more show that this two width image is more similar close to 0.Here set test sample book as X, to be the eigenwert of Y. test sample book be master sample: (x 1, x 2, x 3.., x n), the eigenwert of master sample is (y 1, y 2, y 3..., y n), wherein eigenwert all normalizes in the scope of (0.0,1.0).
For color feature value and Gabor characteristic, value all in (0,1) scope itself, we do not do normalization here.
Discriminant function for each feature is:
E ( x , y ) = Σ k = 0 n ( x k - y k ) 2 , k = 1,2 , . . . , n - - - ( 7 )
For the selection of sorter, if special vehicle instrument type is fewer, the training sample of the instrument of every type is very many, and neural network classifier can be adopted to reach reasonable recognition effect.If the type of special vehicle instrument only has 2 classes, there is a large amount of training samples, the sorter of SVM can be adopted to reach reasonable recognition effect.If the kind of special vehicle instrument type is many, have thousands of kind, can adopt Gaussian modeling or adopt multi-template matching method to carry out classifying and identifying, such recognition effect is quick and accurate.
In the present invention, owing to not being a lot (mainly for 10 kinds of special vehicle instrument type for special vehicle instrument type, as shown in Fig. 3 to Figure 12), and we are to often kind of meter type well-chosen training sample, the quantity of the training sample of often kind of meter type is not very large (for often kind of meter type, the each scale of pointed gets a training sample), all well-chosen representational, therefore we adopt arest neighbors rule to carry out type matching, adopt arest neighbors rule to have extraordinary recognition effect for our this problem.The training sample the most contiguous with test sample book (realtime graphic of input) is found, the training sample Y that is square error is minimum by the most recent method, namely:
Y = min [ Σ k = 0 n ( x k - y k ) 2 ] , k = 1,2 , . . . , n - - - ( 8 )
If Y belongs to certain meter type M, then the Instrument image of test sample book (realtime graphic of input) also belongs to this meter type M.
4, gauge pointer point of fixity, calibration information is obtained according to specific meter type
Relative to industrial instrument, the type identification of special vehicle instrument is necessary more, and feasible.First, special vehicle instrument type is limited, therefore often kind of special vehicle instrument can carry out feature modeling.How a lot of the type of industrial instrument is then, may thousands of kind.Relative industrial instrument, the dial plate of the instrument that special vehicle is dissimilar takes on a different character, and the feature of industrial instrument is once regular, and the scale of a lot of special vehicle instrument neither be uniform, therefore be necessary very much to carry out feature modeling to often kind of dissimilar special vehicle instrument, and to often kind of special vehicle instrument carry out artificial setup parameter value such as its pointer point of fixity, start index and termination graduation position, scale respectively.As Fig. 3 to Figure 12, illustrate pointer point of fixity and the scale distribution parameter value of the special vehicle instrument type setting that the present invention relates to.After identifying the type of special vehicle instrument like this, just can infer according to the setting value of often kind of special vehicle instrument and pointer point of fixity, the important informations such as scale distribution, if pointer just can carry out accurate Meter recognition after automatically identifying further.
Contrast prior art, tool of the present invention has the following advantages:
(1) the present invention extracts characteristics of image, has carried out modeling and type identification to special vehicle instrument type.
(2) color characteristic is for coloured instrument, as dial plate having the meter type effect of red color, green color clearly.
(3) after the present invention identifies meter type automatically, make the pointer point of fixity of special vehicle instrument, the automatic identification of non-uniform scale all becomes possibility.
5, accompanying drawing explanation
Fig. 1. based on the automatic identification process figure of the special vehicle instrument type of characteristics of image.
Fig. 2. based on the modeling process of the special vehicle instrument training set of characteristics of image.
Fig. 3. determine pointer point of fixity, scale, start index and termination scale according to meter type 1.
Fig. 4. determine pointer point of fixity, scale, start index and termination scale according to meter type 2.
Fig. 5. determine pointer point of fixity, scale, start index and termination scale according to meter type 3.
Fig. 6. determine pointer point of fixity, scale, start index and termination scale according to meter type 4.
Fig. 7. determine pointer point of fixity, scale, start index and termination scale according to meter type 5.
Fig. 8. determine pointer point of fixity, scale, start index and termination scale according to meter type 6.
Fig. 9. determine pointer point of fixity, scale, start index and termination scale according to meter type 7.
Figure 10. determine pointer point of fixity, scale, start index and termination scale according to meter type 8.
Figure 11. determine pointer point of fixity, scale, start index and termination scale according to meter type 9.
Figure 12. determine pointer point of fixity, scale, start index and termination scale according to meter type 10.
Embodiment
Below in conjunction with figure, specific embodiment of the invention is further detailed.
The general flow chart of enforcement of the present invention as shown in Figure 1, its flow process is as follows: after the video image utilizing a pair of image capturing system and equipment acquisition special vehicle instrument real-time, be sent to by video data in computing machine, computing machine carries out automatic identifying processing to single width Instrument image.First the disk of this width Instrument image is extracted, then according to disc radius, Instrument image is normalized, color characteristic and Gabor textural characteristics is extracted respectively after normalization, respectively pattern match is carried out to the feature templates in these two kinds of features and training sample, then obtain instrument classification results, finally the recognition result of two kinds is comprehensively obtained meter type.
The acquisition feature templates of training sample of the present invention as shown in Figure 2, to the instrument of every type, the representational training sample of well-chosen, such as each scale of pointed instrument is got a sample image and is placed in training set, then extracts color characteristic and Gabor textural characteristics value for each training sample image.

Claims (4)

1., based on special vehicle instrument type identification and the scaling method of characteristics of image, it is characterized in that, described method is as follows:
First to the special vehicle instrument of every type, gather some special vehicle instrument images, through artificial judgment, representational image is stayed and is used as training sample; For each training sample, by image pre-processing method, picture quality is normalized; Then extract the disk of Instrument image, according to disc radius, size normalization is carried out to Instrument image; Respectively to special vehicle instrument image zooming-out color characteristic and Gabor textural characteristics after normalization; Last for all training samples, respectively with color characteristic and Gabor textural characteristics for vector is every type Modling model;
After the characteristic model of the training sample database of the instrument of every type establishes, for the realtime graphic collected, picture quality normalization is carried out equally through Image semantic classification, extract instrument disk and carry out instrument size normalization, color characteristic and Gabor textural characteristics is extracted respectively after normalization, respectively pattern match is carried out to the feature templates of the instrument training sample of these two kinds of features and each type, obtains instrument classification results by arest neighbors rule;
Described Instrument image normalization is specific as follows:
First pre-service is carried out to special vehicle instrument image, go random noise by medium filtering, being stretched by color strengthens the contrast of Instrument image; Wherein medium filtering adopts the modified adaptive median filter algorithm in conjunction with mean filter and adaptive median filter advantage; And the adaptive neighborhood method of average removal of images interference using band to revise and noise; Wherein, stretched by color and strengthen the contrast of Instrument image: first obtain gray level image corresponding to coloured image, and add up the minimum gray value g1 in Instrument image, gray scale maximal value g2, gray level image is drawn high the gray space of (0,255), make gray level image more clear; If g1 is the minimal gray that total pixel is greater than n, g2 is the maximum gray scale that total pixel is greater than n, described n=10; For any pixel, its gray-scale value is gray (x, y), and the gray-scale value drawing high rear correspondence is GRAY (x, y), then have:
For the coloured image of RGB pattern, the redness of its correspondence, green, blue color draw high and are respectively:
Secondly, the method utilizing Hough transform to detect circular arc obtains the edge of special vehicle instrument disk;
After obtaining the size of instrument disk, linear normalization is carried out to the size of instrument, namely crops the region outside instrument disk, then the pixel in all instrument disks is normalized into the image that radius is 64 pixel sizes; During linear normalization, only do the conversion of equal proportion at X and Y direction; After normalization, Instrument image becomes the rectangle of 128x128 size, and the edge and then instrument disk of rectangular image, the pixel in the region outside instrument disk is all arranged to white pixel.
2. method according to claim 1, is characterized in that, described color feature extracted is specific as follows:
The Instrument image that normalizing is good is divided into the grid of 8X8, each grid adds up total number that is red, green, black picture element respectively, then divided by the area of this grid, constitutes the color feature vector of 8X8X3=192 dimension like this;
For Instrument image based on Gabor gray level image texture feature extraction, for the gray level image of the good instrument of normalization, each gray level image is divided into the grid of 8X8, gets each side's center of a lattice, this generates the sampled point of 8X8; On each sampled point, add a Gaussian wave filter, Gaussian wave filter is here tried to achieve from the Gaussian envelope of Gabor filter, and detailed process as the following formula;
Parameter in the equation above wherein wavelength X=8, N=2 λ;
Like this centered by each sampled point, be that non-zero point is sued for peace by Gaussian wave filter to pixel values all in Gaussian envelope, just on each sampled point of each template image, obtain an eigenwert, altogether obtain sampled point 8X8=64 dimensional feature vector; Proper vector is by obtaining 64 final dimensional feature vectors to the equation of each eigenwert extraction of square root
3. method according to claim 2, is characterized in that, described feature modeling and pattern match specific as follows:
Try to achieve instrument template type image proper vector and by glyphomancy instrument image proper vector after, calculate the Euclidean distance d between them, more show that this two width image is more similar close to 0; Here set test sample book as X, master sample is Y, and the eigenwert of test sample book is: (x 1, x 2, x 3..., x n), the eigenwert of master sample is (y 1, y 2, y 3..., y n), wherein eigenwert all normalizes in the scope of (0.0,1.0);
Discriminant function for each feature is:
For the selection of sorter, adopt arest neighbors rule to carry out type matching, find the training sample the most contiguous with test sample book by nearest neighbor method, described test sample book is the realtime graphic of input, the training sample Y that square error is minimum:
If Y belongs to certain meter type M, then the Instrument image of test sample book also belongs to this meter type M.
4. method according to claim 1, is characterized in that, described method also comprises further:
After identifying the type of special vehicle instrument, infer pointer point of fixity according to the setting value of often kind of special vehicle instrument, the important information of scale distribution, carries out accurate Meter recognition further.
CN201210509245.8A 2012-12-03 2012-12-03 A kind of special vehicle instrument type identification based on characteristics of image and scaling method Expired - Fee Related CN103034838B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210509245.8A CN103034838B (en) 2012-12-03 2012-12-03 A kind of special vehicle instrument type identification based on characteristics of image and scaling method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210509245.8A CN103034838B (en) 2012-12-03 2012-12-03 A kind of special vehicle instrument type identification based on characteristics of image and scaling method

Publications (2)

Publication Number Publication Date
CN103034838A CN103034838A (en) 2013-04-10
CN103034838B true CN103034838B (en) 2015-08-19

Family

ID=48021717

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210509245.8A Expired - Fee Related CN103034838B (en) 2012-12-03 2012-12-03 A kind of special vehicle instrument type identification based on characteristics of image and scaling method

Country Status (1)

Country Link
CN (1) CN103034838B (en)

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103955681A (en) * 2014-05-22 2014-07-30 苏州大学 Human face identification method and system
CN105678344A (en) * 2016-02-29 2016-06-15 浙江群力电气有限公司 Intelligent classification method for power instrument equipment
CN106643965B (en) * 2016-12-23 2020-08-04 贵州电网有限责任公司电力科学研究院 Method for accurately identifying liquid level by utilizing template matching
CN107038049A (en) * 2017-04-25 2017-08-11 航天科技控股集团股份有限公司 A kind of adaptive download system of automobile instrument aging procedure and method based on graphic correlation pattern
CN107122210A (en) * 2017-04-25 2017-09-01 航天科技控股集团股份有限公司 A kind of adaptive download system of automobile instrument program and method based on graphic correlation pattern
CN107291577A (en) * 2017-06-19 2017-10-24 腾讯科技（深圳）有限公司 file backup method, device and storage medium
CN107392198A (en) * 2017-08-03 2017-11-24 北京羿娲科技有限公司 Intelligent meter reading device and system
CN107729906B (en) * 2017-10-24 2021-11-02 国网江苏省电力公司南京供电公司 Intelligent robot-based inspection point ammeter numerical value identification method
CN108875739A (en) * 2018-06-13 2018-11-23 深圳市云识科技有限公司 A kind of accurate detecting method of digital displaying meter reading
CN109272000B (en) * 2018-09-10 2022-01-25 南方电网电力科技股份有限公司 Robot instrument alignment and focusing method and device
CN109409385B (en) * 2018-10-16 2021-02-19 南京鑫和汇通电子科技有限公司 Automatic identification method for pointer instrument
CN110196152A (en) * 2019-03-29 2019-09-03 山东建筑大学 The method for diagnosing faults and system of large-scale landscape lamp group based on machine vision
CN110084317B (en) * 2019-05-06 2023-04-21 北京字节跳动网络技术有限公司 Method and device for recognizing images
CN113805240B (en) * 2020-05-28 2023-06-27 同方威视技术股份有限公司 Vehicle inspection method and system
CN112367461B (en) * 2020-10-27 2021-12-17 京东方科技集团股份有限公司 Instrument image sample manufacturing method and system, storage medium and electronic equipment
CN113029220A (en) * 2021-01-28 2021-06-25 开放智能机器（上海）有限公司 State recognition system and method for industrial instrument panel

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101660932A (en) * 2009-06-15 2010-03-03 浙江大学 Automatic calibration method of pointer type automobile meter

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101660932A (en) * 2009-06-15 2010-03-03 浙江大学 Automatic calibration method of pointer type automobile meter

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
《一种采用Gabor小波的纹理特征提取方法》;张刚等;《中国图象图形学报》;20100228;第15卷(第2期);正文第247-254页 *
《三维重建中特征点提取与匹配算法研究》;乔警卫等;《系统仿真学报》;20080930;第20卷;正文第400-403页 *
《基于Gabor 滤波的灰度图像彩色化方法》;梁琦等;《计算机工程与设计》;20091231;第30卷(第19期);正文第4436-4438、4578页 *
《指针式仪表自动检定系统图像识别技术》;孙琳等;《现代电子技术》;20110415;第34卷(第8期);正文第101-104页 *
魏小莉等.《一种基于马尔科夫随机场模型的彩色纹理图像分割》.《武汉大学学报(信息科学版)》.2010,第35卷(第8期),正文第955-958页. *

Also Published As

Publication number Publication date
CN103034838A (en) 2013-04-10

Similar Documents

Publication Publication Date Title
CN103034838B (en) A kind of special vehicle instrument type identification based on characteristics of image and scaling method
CN103164692B (en) A kind of special vehicle instrument automatic identification system based on computer vision and method
CN106295124B (en) The method of a variety of image detecting technique comprehensive analysis gene subgraph likelihood probability amounts
CN115018844B (en) Plastic film quality evaluation method based on artificial intelligence
CN114549522A (en) Textile quality detection method based on target detection
CN106529559A (en) Pointer-type circular multi-dashboard real-time reading identification method
CN111507426B (en) Non-reference image quality grading evaluation method and device based on visual fusion characteristics
CN106610969A (en) Multimodal information-based video content auditing system and method
CN109919960B (en) Image continuous edge detection method based on multi-scale Gabor filter
CN106503638B (en) Image procossing, vehicle color identification method and system for color identification
CN108564092A (en) Sunflower disease recognition method based on SIFT feature extraction algorithm
CN103914708A (en) Food variety detection method and system based on machine vision
Laga et al. Image-based plant stornata phenotyping
CN105320970A (en) Potato disease diagnostic device, diagnostic system and diagnostic method
CN108073940B (en) Method for detecting 3D target example object in unstructured environment
CN111950654B (en) Magic cube color block color reduction method based on SVM classification
CN111046838A (en) Method and device for identifying wetland remote sensing information
CN113052234A (en) Jade classification method based on image features and deep learning technology
Khamael et al. Using adapted JSEG algorithm with fuzzy C mean for segmentation and counting of white blood cell and nucleus images
CN103871084B (en) Indigo printing fabric pattern recognition method
Wang et al. Color edge detection using the normalization anisotropic Gaussian kernel and multichannel fusion
Vandenbroucke et al. Unsupervised color-image segmentation by multicolor space iterative pixel classification
Di et al. The research on the feature extraction of sunflower leaf rust characteristics based on color and texture feature
Yao et al. An adaptive and quantitative rubber nutrient status analyzing system by digital foliar images
Roux et al. Multiscale anisotropic texture unsupervised clustering for photographic paper

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20150819

Termination date: 20181203

CF01 Termination of patent right due to non-payment of annual fee